A clinical and epidemiological longitudinal survey was made from 1985 to 1989 in four villages of a dracunculiasis hyper endemic region of Benin to identify indices usable in evaluation of prevalence or incidence of dracunculiasis. 720 patients were examined. 93.5% presented a typical eruption of one or more worms. Inflammatory, infectious or mechanical complications were observed in 57.5% of cases and induced disabilities in 44.7% of patients. Although retrospective surveys are easier, authors do not recommend such studies. Cross-sectional surveys would be performed at the peak of transmission. The index would be the number of disabled seen in the village at the moment of investigators' passage. This index represents about 20% of annual dracunculosis cases. Longitudinal surveys would be made by village health workers. The best index is the number of worm eruptions corrected by a reducing factor: from 0.4 to 0.67 according to transmission pattern of the region and knowledge of the local transmission model.